This invention relates to an image processing apparatus and method and a medium, and more particularly to an image processing apparatus and method and a medium wherein an image of an imaging object is picked up and converted into image data and the image data is compressed.
A digital video camera is being popularized which compresses picked up moving pictures using a compression method such as the MPEG (Moving Picture Experts Group) compression method and records the compressed moving pictures onto a disk or a semiconductor memory.
In the MPEG method, data of a finite difference between fields-or frames (in the following description, where there is no particular necessity for identification between a field and a frame, the term “field” is used commonly for them) which neighbor each other with respect to time is calculated to achieve temporal data compression. Further, pixel data are converted into DCT (Discrete Cosine Transform) coefficients to achieve spatial data compression.
A digital video camera of the type described above additionally has a hand-shaking correction function for correcting image data against hand-shaking, that is, an inadvertent movement of the digital video camera when an image is picked up by a user. Two different methods are conventionally available for hand-shaking correction. According to one of the methods, hand-shaking is detected by means of a sensor built in the digital video camera, and picked up image pixels obtained by a CCD (Charge Coupled Device) unit are parallelly translated relative to an image based on the detected hand-shaking. According to the other method, signals read out from such a CCD unit are delayed with respect to time.
In such compression of image data as described above, the relationship between the compression ratio and the picture quality depends upon two factors including the information amount (redundancy) of image data before compression and the amount of motion. In particular, in such compression that the picture quality deterioration between image data before compression and decoded data obtained by compression processing of the image data is lower than a predetermined level, for example, where small subjects exhibit quick and random motions, the compression ratio of image data is low.
For example, when an image of a person in front of a fountain is picked up, in order to suppress the picture quality deterioration of the entire image, MPEG compression allocates a comparatively great amount of coded data to random and quick motion of water of the fountain on the background, but uses a comparatively small amount of coded data for the representation of the person. Since a comparatively great amount of data is allocated to image data which is low in redundancy or exhibits quick motion, the picture quality of the person who is the main imaging object is deteriorated. Where image data is compressed such that the picture quality deterioration may be lower than a predetermined level as described above, the conventional digital video camera has a subject to be solved in that a high compression ratio cannot be applied to image data which is low in redundancy or exhibits quick motion.
Further, since hand-shaking correction is performed independently of compression processing of image data, the conventional digital video camera has another subject to be solved in that it has a great apparatus scale and requires a high cost as much.